CN114761732A - Model sharing system, model management device, and control device for air conditioning device - Google Patents
Model sharing system, model management device, and control device for air conditioning device Download PDFInfo
- Publication number
- CN114761732A CN114761732A CN201980102441.3A CN201980102441A CN114761732A CN 114761732 A CN114761732 A CN 114761732A CN 201980102441 A CN201980102441 A CN 201980102441A CN 114761732 A CN114761732 A CN 114761732A
- Authority
- CN
- China
- Prior art keywords
- model
- air conditioner
- heat load
- learned
- air
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000004378 air conditioning Methods 0.000 title claims description 17
- 238000003860 storage Methods 0.000 claims description 29
- 238000004891 communication Methods 0.000 claims description 26
- 230000005540 biological transmission Effects 0.000 claims description 17
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 4
- 238000004519 manufacturing process Methods 0.000 claims description 2
- 208000033748 Device issues Diseases 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 17
- 238000000034 method Methods 0.000 description 13
- 239000003507 refrigerant Substances 0.000 description 12
- 210000002569 neuron Anatomy 0.000 description 3
- 238000001816 cooling Methods 0.000 description 2
- 238000010438 heat treatment Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 210000004027 cell Anatomy 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
- F24F11/63—Electronic processing
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/30—Control or safety arrangements for purposes related to the operation of the system, e.g. for safety or monitoring
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/04—Programme control other than numerical control, i.e. in sequence controllers or logic controllers
- G05B19/042—Programme control other than numerical control, i.e. in sequence controllers or logic controllers using digital processors
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2110/00—Control inputs relating to air properties
- F24F2110/10—Temperature
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2110/00—Control inputs relating to air properties
- F24F2110/10—Temperature
- F24F2110/12—Temperature of the outside air
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2140/00—Control inputs relating to system states
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/26—Pc applications
- G05B2219/2614—HVAC, heating, ventillation, climate control
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Air Conditioning Control Device (AREA)
Abstract
A model sharing system (1) is provided with: a plurality of control devices (11A, 11B) that control the corresponding controlled devices, respectively; and a model management device (10) that stores the learned model in association with the operating state of the controlled device. The control devices (11A, 11B) acquire a learned model corresponding to an operating state that is the same as or similar to the operating state of the corresponding controlled device from the model management device (10), and control the corresponding controlled device using the acquired learned model. The operation state comprises the following steps: at least one of a type of the controlled device, an environment in which the controlled device is set, and a setting of the controlled device.
Description
Technical Field
The present disclosure relates to a model sharing system, a model management device, and a control device for an air conditioner.
Background
A method of performing high-precision control by using machine learning is known. In this method, a model including a control program and control parameters is generated by learning using a past operation history.
The control devices of a plurality of robots described in patent document 1 share a learned model, thereby shortening the time required for learning.
Patent document 1: japanese patent laid-open publication No. 2017-30135
However, when the same learned model is shared among a plurality of control devices, the accuracy of control is low when the control content of the controlled device greatly differs depending on the operating state of the controlled device.
For example, in an air conditioner, the difficulty of cooling and heating rooms greatly varies depending on the material of an object to be installed, the climate, the device structure, and the like. Therefore, in a method in which the same model is shared among the control devices of a plurality of air conditioners, highly accurate control cannot be achieved.
Disclosure of Invention
Therefore, an object of the present disclosure is to provide a model sharing system, a model management device, and a control system for an air conditioning system, which are capable of performing highly accurate control even when a large difference occurs in the control content of a controlled device due to the operating state of the controlled device.
The model sharing system of the present disclosure includes: a plurality of control devices that control the corresponding controlled devices, respectively; and a model management device that stores the learned model in association with the operation state of the controlled device. The control device acquires a learned model corresponding to an operating state that is the same as or similar to the operating state of the corresponding controlled device from the model management device, and controls the corresponding controlled device using the acquired learned model. The operation state comprises the following steps: at least one of a type of the controlled device, an environment in which the controlled device is set, and a setting of the controlled device.
The present disclosure relates to a model management device of a model sharing system for sharing a plurality of learned heat load models among control devices of a plurality of air conditioners. The model management device is provided with: a model storage unit that stores a plurality of learned heat load models in association with an operating state of the air conditioner; a communication unit capable of communicating with a control device of a plurality of air conditioning devices; a model providing unit that provides, to the control device of the air conditioner, a heat load model corresponding to an operation state that is the same as or most similar to the operation state of the air conditioner, from among the plurality of learned heat load models stored in the model storage unit, in response to a transmission request from the control device of the air conditioner to specify the operation state of the air conditioner; and a model registration unit that acquires a learned heat load model that specifies an operating state of the air conditioner from the air conditioner, and stores the acquired learned heat load model in the model storage unit in accordance with the acquired operating state of the air conditioner.
The control device for an air conditioner according to the present disclosure includes a communication unit that can communicate with a model management device that manages a learned heat load model shared among control devices for a plurality of air conditioners. The model management device stores the learned heat load model in association with the operating state of the air conditioner. The control device for an air conditioner according to the present disclosure further includes a control unit that issues a transmission request of a heat load model specifying an operation state of the air conditioner, and acquires a learned heat load model transmitted from the model management device in response to the transmission request. The learned heat load model corresponds to an operating state that is the same as or most similar to the specified operating state.
The control device for an air conditioner according to the present disclosure further includes a learning unit that acquires input data and teacher data of a heat load model for additional learning by operating the air conditioner, and performs additional learning on the acquired heat load model using the acquired input data and teacher data. The control unit controls the air conditioner using the additionally learned heat load model. The communication unit transmits the additionally learned thermal load model to the model management device together with the operating state of the air conditioner. The operation state comprises the following steps: the type of the air conditioner, the environment in which the air conditioner is installed, and the setting of the air conditioner.
According to the present disclosure, the control device acquires a learned model corresponding to an operating state that is the same as or similar to the operating state of the corresponding controlled device from the model management device, and controls the corresponding controlled device using the acquired learned model. Thus, even when the control content of the controlled device has a large difference due to the operating state of the controlled device, it is possible to execute highly accurate control.
Drawings
Fig. 1 is a diagram showing the overall configuration of a model sharing system 1 according to embodiment 1.
Fig. 2 is a diagram showing an example of information stored in the model storage unit 101.
Fig. 3 is a diagram showing an example of the structure of the air conditioner.
Fig. 4 is a diagram for explaining an example of the control of the air conditioner 2A by the control unit 112A.
Fig. 5 is a diagram showing an example of the thermal load model.
Fig. 6 is a diagram showing an example of output data of the heat load model.
Fig. 7 is a diagram showing an example of output data of the heat load model.
Fig. 8 is a diagram showing an example of the operating state.
Fig. 9 is a flowchart showing a procedure in which the control device 11A acquires the heat load model from the model management device 10 immediately after the operation of embodiment 1.
Fig. 10 is a flowchart showing a procedure of performing additional learning by the control device 11A according to embodiment 1.
Fig. 11 is a diagram showing a hardware configuration of the model management device 10 and the control devices 11A and 11B.
Detailed Description
Hereinafter, a model sharing system and the like according to the embodiments will be described with reference to the drawings and the like. In the following drawings, the same or corresponding portions are designated by the same reference numerals and are used in common throughout the embodiments described below. In the drawings, the relationship between the sizes of the respective components may be different from the actual one. The embodiments of the constituent elements shown throughout the specification are merely examples, and are not limited to the embodiments described in the specification. All the devices described in the specification may not be included. In particular, the combination of the components is not limited to the combination in each embodiment, and the components described in other embodiments may be applied to another embodiment. When it is not necessary to particularly distinguish and identify a plurality of devices of the same kind, etc., which are distinguished by subscripts, reference numerals, subscripts, etc. may be omitted from the description.
Fig. 1 is a diagram showing the overall configuration of a model sharing system 1 according to embodiment 1.
The model sharing system 1 shares a plurality of learned heat load models between the control devices 11A and 11B of the plurality of air conditioners 2a and 2B.
The model sharing system 1 includes a model management device 10 and a plurality of control devices 11A and 11B.
The model management device 10 is communicably connected to the control device 11A and the control device 11B via an electrical communication line 13. The model management device 10 and the control devices 11A and 11B can communicate the heat load model.
The model management device 10 includes a communication unit 104, a model providing unit 102, a model registering unit 103, and a model storing unit 101.
The model storage unit 101 stores information indicating the learned heat load model in association with the operating state of the air conditioner.
Fig. 2 is a diagram showing an example of information stored in the model storage unit 101.
The model storage unit 101 stores information indicating the heat load models M (1) to M (n) in association with the operating states S (1) to S (n) of the air conditioning apparatus. When the thermal load models M (1) to M (n) are neural networks, the model storage unit 101 stores weighting coefficients of the neural networks as information indicating the thermal load models M (1) to M (n).
The communication unit 104 is configured to be able to communicate with the control devices 11A and 11B via the electric communication line 13.
The model providing unit 102 receives a request for transmission of a heat load model from any one of the control devices 11A and 11B of the plurality of air conditioners.
When the control device 11A is the control device that has transmitted the transmission request, the model providing unit 102 provides the control device 11A of the air conditioner with a heat load model corresponding to an operation state that is the same as or most similar to the operation state of the air conditioner 2a among the plurality of learned heat load models stored in the model storage unit 101. For example, the model providing unit 102 normalizes the two operation states by normalizing the value of each item representing the operation state of the air conditioner 2a and the value of each item representing the operation state stored in the model storage unit 101. The model providing unit 102 calculates the similarity between the operating state of the air conditioner 2a and the operating state stored in the model storage unit 101 based on the euclidean distance between the two operating states after normalization.
When the control device 11B is the control device that has transmitted the transmission request, the model providing unit 102 provides the control device 11B of the air conditioner with the heat load model corresponding to the operation state that is the same as or most similar to the operation state of the air conditioner 2B among the plurality of learned heat load models stored in the model storage unit 101. For example, the model providing unit 102 normalizes the two operation states by normalizing the value of each item representing the operation state of the air conditioner 2b and the value of each item representing the operation state stored in the model storage unit 101. The model providing unit 102 calculates the similarity between the operating state of the air conditioner 2b and the operating state stored in the model storage unit 101 based on the euclidean distance between the two operating states after normalization.
The above-described method of calculating the similarity is an example. The method of calculating the similarity may be any method as long as it is a method of calculating based on the operation state. For example, the normalized values may be weighted.
The model registration unit 103 acquires a set of the learned heat load model and the operating state of the air conditioner transmitted from either of the control devices 11A and 11B. The model registration unit 103 stores the acquired learned thermal load model in the model storage unit 101 in association with the acquired operating state of the air conditioner.
The control device 11A includes: a communication unit 114A, a learning unit 113A, a model storage unit 110A, a control unit 112A, and an operation state collection unit 111A.
The communication unit 114A is configured to be able to communicate with the model management device 10 via the electric communication line 13.
The model storage unit 110A stores a learned heat load model obtained by additionally learning the learned heat load model obtained from the model management device 10 or the learned heat load model obtained from the model management device 10.
The operation state collection unit 111A collects information on the operation state of the air conditioner 2 a.
The control unit 112A issues a transmission request of a heat load model in which the operating state of the air conditioner 2A is specified. The control unit 112A acquires the learned thermal load model transmitted from the model management device 10 in response to the transmission request of the thermal load model, and stores the acquired thermal load model in the model storage unit 110A. The learned heat load model transmitted from the model management device 10 corresponds to an operation state that is the same as or most similar to the operation state of the air conditioner 2a included in the transmission request.
The learning unit 113A operates the air conditioner 2a to acquire input data of a heat load model for additional learning and teacher data. The learning unit 113A additionally learns the heat load model stored in the model storage unit 110A using the acquired input data and teacher data.
The control unit 112A controls the air conditioner 2A using the additionally learned heat load model. The communication unit 115A executes a control command from the control unit 112A to the air conditioner 2A and communication of sensor data from the air conditioner 2A to the control unit 112A.
The communication unit 114A transmits the additionally learned heat load model to the model management device 10 together with the operating state of the air conditioner 2 a.
The configuration of the control device 11B is also the same as that of the control device 11A, and therefore, description thereof will not be repeated.
Fig. 3 is a diagram showing an example of the configuration of the air conditioner 2 a.
The air conditioner 2a includes an outdoor unit 50 and a plurality of indoor units 40a and 40b.
The outdoor unit 50 includes: a compressor 51 that compresses and discharges a refrigerant; a heat source side heat exchanger 52 that exchanges heat between the outside air and the refrigerant; and a four-way valve 53 that switches the direction of refrigerant flow according to the operation mode. The outdoor unit 50 includes an outside air temperature sensor 54 for detecting the outside air temperature.
The indoor unit 40a includes: a load side heat exchanger 41a that exchanges heat between indoor air and refrigerant, and an expansion device 42a that decompresses and expands high-pressure refrigerant. The indoor unit 40a includes an indoor temperature sensor 43a for detecting a room temperature.
The indoor unit 40b includes: a load side heat exchanger 41b that exchanges heat between indoor air and refrigerant, and an expansion device 42b that decompresses and expands high-pressure refrigerant. The indoor unit 40b includes an indoor temperature sensor 43b that detects a room temperature.
The compressor 51 is, for example, an inverter type compressor capable of changing capacity by changing an operating frequency. The expansion devices 42a and 42b are, for example, electronic expansion valves.
In the outdoor unit 50 and the indoor unit 40a, the compressor 51, the heat source side heat exchanger 52, the expansion device 42a, and the load side heat exchanger 41a are connected to constitute a refrigerant circuit 60 in which a refrigerant circulates. In the outdoor unit 50 and the indoor unit 40b, the compressor 51, the heat source side heat exchanger 52, the expansion device 42b, and the load side heat exchanger 41b are connected to constitute a refrigerant circuit 60 in which a refrigerant circulates.
The configuration of the air conditioner 2b is the same as that of the air conditioner 2a, and therefore, description thereof will not be repeated.
Fig. 4 is a diagram for explaining an example of the control of the air conditioner 2A by the control unit 112A.
When the indoor unit 40a is operated, the control unit 112A controls the operating frequency of the compressor 51 and the opening degree of the expansion device 42A based on the outside air temperature detected by the outside air temperature sensor 54, the room temperature detected by the indoor temperature sensor 43a, and the set temperature. When the indoor unit 40b is operated, the control unit 112A controls the operating frequency of the compressor 51 and the opening degree of the expansion device 42b based on the outside air temperature detected by the outside air temperature sensor 54, the room temperature detected by the indoor temperature sensor 43b, and the set temperature.
When the indoor units 40a and 40b are operated, the control unit 112A controls the operating frequency of the compressor 51 and the opening degrees of the expansion devices 42A and 42b based on the outside air temperature detected by the outside air temperature sensor 54, the room temperature and the set temperature of the indoor unit 40a, and the room temperature and the set temperature of the indoor unit 40b.
The control unit 112A switches the flow path of the four-way valve 53 between when the operation mode of the air conditioner is the cooling operation mode and when the operation mode is the heating operation mode.
The control unit 112A controls additional learning of the learned heat load model stored in the model storage unit 110A. The control unit 112A controls the air conditioner 2A during operation using the learned heat load model stored in the model storage unit 110A.
The control of the air conditioner 2B by the control unit 112B is the same as the control of the air conditioner 2A by the control unit 112A, and therefore, the description thereof will not be repeated.
The learning unit 113A generates a thermal load model using supervised learning using learning data. The learning unit 113A corrects the thermal load model (additional learning) by using supervised learning using additional learning data. Supervised learning refers to learning features in learning data composed of input and results (labels) by giving a large number of sets of the learning data to a learning unit. This enables the estimation result to be input (generalization).
Fig. 5 is a diagram showing an example of the heat load model.
As shown in fig. 5, the thermal load model is composed of a neural network. The neural network is composed of the following layers: an input layer composed of a plurality of neurons, an intermediate layer (hidden layer) composed of a plurality of neurons, and an output layer composed of a plurality of neurons. The intermediate layer may be one or more than two layers. Input data x (i) is given to the i-th cell of the input layer. The output data Z is output from the output layer.
The input data X (1) to X (n) are data indicating factors affecting the heat load of the air conditioner 2. The output data Z is data indicating the heat load of the air conditioner 2.
Fig. 6 is a diagram showing an example of input data of the thermal load model.
As shown in fig. 6, the input data of the thermal load model includes: the difference between the set temperature and the outside air temperature, the difference between the set temperature and the indoor temperature, and the frequency of the compressor provided in the air conditioner 2.
Fig. 7 is a diagram showing an example of output data of the heat load model.
As shown in fig. 7, the output data Z of the heat load model is the time from the start of the operation of the indoor unit 40 until the indoor temperature reaches the set temperature.
Fig. 8 is a diagram showing an example of the operating state.
The operation state comprises the following steps: the type of the air conditioner 2, the environment in which the air conditioner 2 is installed, and the settings of the air conditioner 2.
The types of air conditioners 2 include: at least one of the number of outdoor units 50 of the air-conditioning apparatus 2, the number of indoor units 40 of the air-conditioning apparatus 2, and the manufacturing number of the air-conditioning apparatus 2.
The environment in which the air conditioner 2 is installed includes: at least one of a place where the air conditioner 2 is installed and a size of a room where the air conditioner 2 is installed.
The setting of the air conditioner 2 includes the amount of change in the indoor temperature over a certain period of time due to the operation of the air conditioner 2.
The control unit 112A acquires the heat load model from the model management device 10 based on the operating state of the air conditioner 2A. The learning unit 113A additionally learns the acquired heat load model using the learning data obtained at the time of the test operation.
The control unit 112A gives input data to the additionally learned heat load model during operation of the air conditioner 2A, and acquires output data of the additionally learned heat load model. The input data is at least one of a difference between the set temperature and the outside air temperature, a difference between the set temperature and the indoor temperature, and a frequency of a compressor provided in the air conditioner. The output data is the time from the start of the operation of the indoor unit 40 until the indoor temperature reaches the set temperature. For example, the control unit 112A determines a schedule such as an operation start time of the air conditioner 2A based on the output data.
Fig. 9 is a flowchart showing a procedure in which the control device 11A acquires the heat load model from the model management device 10 immediately after the operation of embodiment 1.
In step S101, the operation state collection unit 111A of the control device 11A acquires the operation state of the air conditioner 2 a. In the air conditioner 2a, since the indoor temperature, the number of outdoor units 50, and the number of indoor units 40, which change over a certain period of time, greatly affect the control, the operation state collection unit 111A acquires these pieces of information.
The values of the respective items of the operating state are set to an upper limit value and a lower limit value. When the value of each item of the acquired operation state of the air conditioner 2a exceeds the upper limit value, the operation state collection unit 111A lowers the value of each item to the upper limit value. When the value of each item of the acquired operation state of the air conditioner 2a is lower than the lower limit value, the operation state collection unit 111A increases the value of each item to the lower limit value.
In step S102, the control unit 112A of the control device 11A transmits a transmission request specifying the learned heat load model of the operating state acquired in step S101.
In step S103, the communication unit 114A of the control device 11A transmits a transmission request for specifying the learned heat load model of the operation state issued in step S102 to the model management device 10.
In step S104, the communication unit 104 of the model management device 10 receives a request for transmitting the learned thermal load model in which the operating state is specified.
In step S105, the model providing unit 102 of the model management device 10 outputs the learned heat load model corresponding to the operation state that is the same as or most similar to the specified operation state among the heat load models stored in the model storage unit 101 to the communication unit 104.
In step S106, the communication unit 104 of the model management device 10 transmits the learned heat load model output from the model providing unit 102 to the control device 11A that has issued the transmission request.
In step S107, the communication unit 114A of the control device 11A receives the learned heat load model.
In step S108, the control unit 112A of the control device 11A stores the received learned heat load model in the model storage unit 110A.
The procedure for the control device 11B to obtain the heat load model from the model management device 10 immediately after the operation is the same as that in fig. 9, and therefore, the description thereof will not be repeated.
Fig. 10 is a flowchart showing a procedure of performing additional learning by the control device 11A according to embodiment 1.
In step S201, the control unit 112A of the control device 11A performs a test operation of the air conditioner 2A to acquire additional learning data for learning, which is composed of input data and teacher data.
In step S202, the control unit 112A of the control device 11A reads the heat load model stored in the model storage unit 110A. The control unit 112A additionally learns the read thermal load model using the acquired learning data for additional learning.
In step S203, the operation state collection unit 111A of the control device 11A acquires the operation state of the air conditioner 2 a. The operation state collection unit 111A acquires information on, for example, the indoor temperature, the number of outdoor units 50, and the number of indoor units 40 that change over a certain period of time.
In step S204, the control unit 112A of the control device 11A issues a registration request including the operating state acquired in step S203 and the additionally learned thermal load model.
In step S205, the communication unit 114A of the control device 11A transmits the registration request issued in step S204 to the model management device 10.
In step S206, the communication unit 104 of the model management device 10 receives a registration request including the operation state and the additionally learned thermal load model.
In step S207, the model registration unit 103 of the model management device 10 stores the additionally learned heat load model included in the registration request in the model storage unit 101 in association with the operation state included in the registration request.
After step S205, the following is executed in the control device 11A.
The procedure of performing the additional learning by the control device 11B is the same as that in fig. 10, and therefore, the description thereof will not be repeated.
The present disclosure is not limited to the above embodiments. For example, the following modifications are also included.
(1) In the model management device 10 and the control devices 11A and 11B described in the above embodiments, when the functions of the model management device 10 and the control devices 11A and 11B, which can be implemented by hardware or software of a digital circuit, are realized by software, the model management device 10 and the control devices 11A and 11B include, for example, a processor 5002 and a memory 5001 connected by a bus 5003 as shown in fig. 11, and the processor 5002 can execute a program stored in the memory 5001. The processor 5002 includes a main processor, a communication processor, and the like. The memory 5001 is constituted by a RAM, a flash memory, a hard disk, or the like.
(2) In the above-described embodiment, the heat load model of the air conditioner is described as an example of the learned model shared among the plurality of control devices, but the present invention is not limited to this. The shared learning-completed model may be any model as long as it is a model for controlling the controlled device. In this case, the operation state includes at least one of the kind of the controlled device, the environment set by the controlled device, and the setting of the controlled device.
(3) The model management apparatus 10 may also exist on a cloud server.
(4) In the above-described embodiment, the case where the neural network is applied as the learning algorithm of the thermal load model has been described, but the present invention is not limited to this. Other machine learning algorithms such as support vector machines may also be used.
(5) In the above-described embodiment, the same heat load model is used for the input data items and the output data items even if the operating state is different, but the present invention is not limited to this. Depending on the operating state, a thermal load model may be used in which the items of input data are different from those of output data.
(6) The sequence of the flowchart of fig. 10 may be performed daily. The operation schedule of the air conditioner does not need to be decided on the day when the air conditioner is not used, and therefore the sequence of the flowchart of fig. 10 may not be executed.
The embodiments disclosed herein are to be considered in all respects as illustrative and not restrictive. The scope of the present disclosure is defined by the claims rather than the description above, and is intended to include meanings equivalent to the claims and all modifications within the scope.
Description of the reference numerals
A model sharing system; 2a, 2b. A model management device; 11A, 11b.. the control device; an electrical communication line; an indoor unit 40a, 40b.. 9; 41a, 41b.. load side heat exchangers; 42a, 42b.. the expansion device; 43a, 43b.. indoor temperature sensor; an outdoor unit; a compressor; a heat source side heat exchanger; 53.. a four-way valve; an outside air temperature sensor; a refrigerant circuit; 101. a model storage portion; a model providing section; a model registration section; 104. 114A, 115A, 114B, 115B.. the communication section; 111A, 111b.. the operation state collecting section; 112A, 112b.. the control section; 113A, 113b.. the learning section; a memory; a processor; a bus.
Claims (13)
1. A model sharing system is characterized by comprising:
a plurality of control devices that control the corresponding controlled devices, respectively; and
a model management device that stores a learned model in correspondence with an operation state of the controlled device,
the control device acquires a learned model corresponding to an operating state that is the same as or similar to the operating state of the corresponding controlled device from the model management device, and controls the corresponding controlled device using the acquired learned model,
the operating states include: at least one of a type of the controlled device, an environment in which the controlled device is set, and a setting of the controlled device.
2. Model sharing system according to claim 1,
the controlled device is an air-conditioning device,
the model is a thermal load model of the air conditioning unit,
the input data of the thermal load model is data representing factors that influence the thermal load of the air conditioner,
the output data of the heat load model is data representing the heat load of the air conditioner.
3. Model sharing system according to claim 2,
the input data includes: at least one of a difference between the set temperature and the outside air temperature, a difference between the set temperature and the indoor temperature, and a frequency of a compressor provided in the air conditioning apparatus.
4. Model sharing system according to claim 2,
the output data is a time from the start of operation of an indoor unit of the air conditioner to a change in the indoor temperature to a set temperature.
5. Model sharing system according to claim 2,
in the operating state, the air conditioning device includes, as types: at least one of the number of outdoor units of the air conditioner, the number of indoor units of the air conditioner, and a manufacturing number of the air conditioner.
6. Model sharing system according to claim 2,
in the operating state, the environment as the environment in which the air conditioner is disposed includes: at least one of a place where the air conditioner is installed and a size of a room where the air conditioner is installed.
7. Model sharing system according to claim 2,
in the operating state, the setting as the air conditioner includes: a change amount of an indoor temperature within a certain time due to an operation of the air conditioner.
8. Model sharing system according to any of claims 2 to 7,
the control device issues a transmission request of a heat load model in which an operation state of the air conditioner is designated, and acquires a learned heat load model transmitted from the model management device in accordance with the transmission request, the acquired learned heat load model corresponding to an operation state that is the same as or most similar to the designated operation state.
9. Model sharing system according to any of claims 2 to 8,
the control device acquires input data and teacher data of the heat load model for additional learning by operating the corresponding air conditioner, and performs additional learning on the acquired heat load model using the acquired input data and teacher data.
10. Model sharing system according to claim 9,
the control device transmits the additionally learned heat load model to the model management device together with the operating state,
the model management device stores the received additionally learned heat load model in association with the received operating state.
11. Model sharing system according to any of claims 2 to 10,
the thermal load model is composed of a neural network.
12. A model management device of a model sharing system for sharing a plurality of learned heat load models among control devices of a plurality of air conditioners, comprising:
a model storage unit that stores a plurality of learned heat load models in association with an operation state of the air conditioner;
a communication unit capable of communicating with the control devices of the plurality of air conditioning devices;
a model providing unit that provides, to a control device of the air-conditioning apparatus, a heat load model corresponding to an operation state that is the same as or most similar to an operation state of the air-conditioning apparatus, from among a plurality of learned heat load models stored in the model storage unit, in response to a transmission request from the control device of the air-conditioning apparatus to specify the operation state of the air-conditioning apparatus; and
and a model registration unit that acquires the learned thermal load model, which specifies the operating state of the air conditioner, from the air conditioner, and stores the acquired learned thermal load model in the model storage unit in association with the acquired operating state of the air conditioner.
13. A control device for an air conditioner, characterized in that,
the air conditioner control device includes a communication unit capable of communicating with a model management device that manages a learned heat load model shared among control devices of a plurality of air conditioners, the model management device storing the learned heat load model in association with an operation state of the air conditioners,
the control device for an air conditioner further includes a control unit that issues a transmission request of a heat load model that specifies an operating state of the air conditioner, and acquires a learned heat load model transmitted from the model management device in accordance with the transmission request, the learned heat load model corresponding to an operating state that is the same as or most similar to the specified operating state,
the control device of the air conditioner further comprises a learning unit that acquires input data and teacher data of the thermal load model for additional learning by operating the air conditioner, and performs additional learning on the acquired thermal load model by using the acquired input data and teacher data,
the control unit controls the air conditioner using the additionally learned heat load model,
the communication unit transmits the additionally learned heat load model to the model management device together with the operating state of the air conditioner,
the operating states include: at least one of a type of the air conditioner, an environment in which the air conditioner is installed, and a setting of the air conditioner.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2019/048993 WO2021117234A1 (en) | 2019-12-13 | 2019-12-13 | Model sharing system, model management apparatus, and control apparatus for air conditioning apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114761732A true CN114761732A (en) | 2022-07-15 |
| CN114761732B CN114761732B (en) | 2024-03-19 |
Family
ID=76330123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201980102441.3A Active CN114761732B (en) | 2019-12-13 | 2019-12-13 | Model sharing system, model management device, and control device for air conditioner |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US20220333810A1 (en) |
| JP (1) | JP7378497B2 (en) |
| CN (1) | CN114761732B (en) |
| DE (1) | DE112019007970T5 (en) |
| WO (1) | WO2021117234A1 (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210396404A1 (en) * | 2020-06-22 | 2021-12-23 | Micah Laughmiller | Innovative System for Providing Hyper Efficient HVAC |
| JP2024031381A (en) * | 2022-08-26 | 2024-03-07 | 三菱重工サーマルシステムズ株式会社 | Control device, control method and air conditioner |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0886490A (en) * | 1994-09-14 | 1996-04-02 | Toshiba Corp | Predicting equipment of thermal load |
| JPH08275265A (en) * | 1995-04-04 | 1996-10-18 | Tokyo Gas Co Ltd | Fault diagnostic device |
| JP2007240067A (en) * | 2006-03-09 | 2007-09-20 | Hitachi Ltd | Air-conditioning control system |
| JP2010055303A (en) * | 2008-08-27 | 2010-03-11 | Denso It Laboratory Inc | Learning data management device, learning data management method and air-conditioner for vehicle, and control device of apparatus |
| JP2014203300A (en) * | 2013-04-05 | 2014-10-27 | キヤノン株式会社 | Content management device, content management method, and program |
| WO2015173842A1 (en) * | 2014-05-12 | 2015-11-19 | 三菱電機株式会社 | Parameter learning device and parameter learning method |
| WO2015174176A1 (en) * | 2014-05-12 | 2015-11-19 | 三菱電機株式会社 | Ventilation controller and method for controlling ventilation |
| US20160161137A1 (en) * | 2014-12-04 | 2016-06-09 | Delta Electronics, Inc. | Controlling system for environmental comfort degree and controlling method of the controlling system |
| US20170307246A1 (en) * | 2015-01-22 | 2017-10-26 | Johnson Controls-Hitachi Air Condictioning Technology (Honh Kong) Limited | Maintenance system and maintenance method of air conditioner |
| JP2018071853A (en) * | 2016-10-27 | 2018-05-10 | インフォグリーン株式会社 | Learning device, control device, learning method, control method, learning program, and control program |
| CN109612037A (en) * | 2017-10-04 | 2019-04-12 | 发那科株式会社 | Air-conditioner control system |
-
2019
- 2019-12-13 CN CN201980102441.3A patent/CN114761732B/en active Active
- 2019-12-13 JP JP2021563571A patent/JP7378497B2/en active Active
- 2019-12-13 US US17/760,851 patent/US20220333810A1/en active Pending
- 2019-12-13 DE DE112019007970.0T patent/DE112019007970T5/en active Pending
- 2019-12-13 WO PCT/JP2019/048993 patent/WO2021117234A1/en active Application Filing
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0886490A (en) * | 1994-09-14 | 1996-04-02 | Toshiba Corp | Predicting equipment of thermal load |
| JPH08275265A (en) * | 1995-04-04 | 1996-10-18 | Tokyo Gas Co Ltd | Fault diagnostic device |
| JP2007240067A (en) * | 2006-03-09 | 2007-09-20 | Hitachi Ltd | Air-conditioning control system |
| JP2010055303A (en) * | 2008-08-27 | 2010-03-11 | Denso It Laboratory Inc | Learning data management device, learning data management method and air-conditioner for vehicle, and control device of apparatus |
| JP2014203300A (en) * | 2013-04-05 | 2014-10-27 | キヤノン株式会社 | Content management device, content management method, and program |
| WO2015173842A1 (en) * | 2014-05-12 | 2015-11-19 | 三菱電機株式会社 | Parameter learning device and parameter learning method |
| WO2015174176A1 (en) * | 2014-05-12 | 2015-11-19 | 三菱電機株式会社 | Ventilation controller and method for controlling ventilation |
| US20160161137A1 (en) * | 2014-12-04 | 2016-06-09 | Delta Electronics, Inc. | Controlling system for environmental comfort degree and controlling method of the controlling system |
| US20170307246A1 (en) * | 2015-01-22 | 2017-10-26 | Johnson Controls-Hitachi Air Condictioning Technology (Honh Kong) Limited | Maintenance system and maintenance method of air conditioner |
| JP2018071853A (en) * | 2016-10-27 | 2018-05-10 | インフォグリーン株式会社 | Learning device, control device, learning method, control method, learning program, and control program |
| CN109612037A (en) * | 2017-10-04 | 2019-04-12 | 发那科株式会社 | Air-conditioner control system |
| JP2019066135A (en) * | 2017-10-04 | 2019-04-25 | ファナック株式会社 | Air-conditioning control system |
Also Published As
| Publication number | Publication date |
|---|---|
| US20220333810A1 (en) | 2022-10-20 |
| JPWO2021117234A1 (en) | 2021-06-17 |
| DE112019007970T5 (en) | 2022-09-22 |
| CN114761732B (en) | 2024-03-19 |
| JP7378497B2 (en) | 2023-11-13 |
| WO2021117234A1 (en) | 2021-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11774923B2 (en) | Augmented deep learning using combined regression and artificial neural network modeling | |
| US10747187B2 (en) | Building management system with voting-based fault detection and diagnostics | |
| CN112577161B (en) | Air conditioner energy consumption model training method and air conditioner system control method | |
| Afram et al. | Review of modeling methods for HVAC systems | |
| US11085663B2 (en) | Building management system with triggered feedback set-point signal for persistent excitation | |
| US20230312174A1 (en) | Variable refrigerant flow system with zone grouping | |
| CA3040117C (en) | Operating an hvac system based on predicted indoor air temperature | |
| CN107120794B (en) | Air conditioner operation condition adjusting method and air conditioner | |
| CN114761732B (en) | Model sharing system, model management device, and control device for air conditioner | |
| US11236917B2 (en) | Building control system with zone grouping based on predictive models | |
| US20210018204A1 (en) | Variable refrigerant flow system with zone grouping control feasibility estimation | |
| JP2005301582A (en) | Process management device | |
| CN115176115B (en) | Heat load estimation device, air conditioner control system, and heat load estimation method | |
| CN113641163B (en) | Self-configuration method and system of environment control equipment, computer equipment and storage medium | |
| US20210192469A1 (en) | Building control system with peer analysis for predictive models | |
| JP2020139705A (en) | Operation control method, operation control program and operation control device | |
| CN116097046B (en) | Generating method, information processing apparatus, information processing method, and learning-completed model | |
| US11768003B2 (en) | Variable refrigerant flow system with zone grouping | |
| WO2024114675A1 (en) | Multi-split air conditioning system and control method thereof | |
| WO2023197711A1 (en) | Multi-split air-conditioning system, fault positioning method, and fault diagnosis model training method | |
| KR20180068025A (en) | System for controlling Air-Conditioner | |
| CN115997221A (en) | Generating method, program, information processing device, information processing method, and learning-completed model | |
| CN118049727A (en) | Abnormality detection method of air conditioner and electronic equipment | |
| CN117648740A (en) | Central air conditioner modeling method | |
| CN113757915A (en) | Abnormality diagnosis device and abnormality diagnosis method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |